LAUSR

Abstract This paper addresses the problem of predicting thermal diffusion ( α ) and conductivity ( k ) during a microwave heating process. The estimation required the use of “simulated thermal images” for the microwave heating process of a solid parallelepiped sample made of SiC, and the application of the Peak Signal-to-Noise Ratio (PSNR) criterion. White Gaussian noise was also incorporated into the problem to simulate a realistic situation. The inverse problem was solved by applying three global optimization methods such as the Spiral Optimization Algorithm (SOA), the Vortex Search (VS) algorithm, and the Weighted Attraction Method (WAM). Results show that all algorithms converge into the expected solution and, therefore, were completely acceptable. However, the VS algorithm was the most accurate, improving the iterations and the parameters for function evaluation in nearly 2.5 times with respect to SOA and two times with regard to the WAM algorithm.